Overconfidence and risk behavior in family firms

Abstract

This study proposes an original theoretical contribution on the risk behavior of family firms. Traditionally, individual decision making is deemed to have a strong influence on the risk behavior of an organization, in the sense that firm’s choices are biased by individual characteristics. In non-family firms, the responsibility for the risk behavior of the company only depends on managers. By contrast, in family businesses there is an interplay between biases of those family members actually controlling the company and top managers choices. Whilst the agency theory offers an effective explanation of the relationship between the agent and the principal, the family business stream of research still lacks of such a theory, able to explain the impact of the principal behavior on the organization’s risk propensity. Our study aims to tackle this huge and relevant gap. We study how the risk propensity of family firms is influenced by those family members who are in direct control of the company (e.g. by participating of the board of directors). We focus on a specific behavioral characteristic, which is overconfidence. Overconfidence roughly occurs as the overestimation of the precision of the individual knowledge, both in the negative and the positive domain. As the consequence, the future is seen as either fulfilled with hopes or “Corporate Governance: Search for the Advanced Practices” Rome, February 28, 2019 29 the opposite, alternatively. After extensively analyzing the literature on the topic, we propose a conceptual model for the interplay between overconfidence and risk behavior in family firms. We hypothesize that belief of those family members who exert control over the firms are a relevant predictor of the risk behavior of the company. In particular, we focus on how the overconfidence of family members predicts the risk propensity in family firms. Our theoretical contribution brings substantial knowledge advancement in studies on family firms by filling a resonating gap, which demands future empirical testing.